An Integrative Approach to Assess and Map Zostera noltei Meadows Along the Romanian Black Sea Coast

Abstract

Seagrass meadows, including those formed by Zostera noltei, play a crucial role in marine ecosystem health by providing habitat stability and coastal protection. In the Romanian Black Sea, Z. noltei meadows are critically endangered due to pressures from eutrophication, habitat loss, and climate change. This study presents a comprehensive baseline assessment of Z. noltei meadows near Mangalia, Romania, utilizing in situ field methods and UAV mapping conducted in the spring and summer of 2023. Seven meadow sites (Z1–Z7) were identified, with notable variability in density, shoot counts, and coverage across sites. Site Z1 exhibited the highest density (1223 shoots/m⁻²) and Z5 and Z7 the longest leaves (an average of 60 cm), reflecting possible environmental influences. Statistical analyses revealed significant inter-site differences in shoot density and leaf length, with density emerging as a primary differentiator. Ex situ analyses of epiphyte load indicated a median, balanced epiphyte load. This baseline dataset supported the selection of Z1 as a reference donor site for seagrass relocation activities along the Romanian coast in 2023. By providing critical insights into Z. noltei structure and health, this study supports future conservation efforts and evidence-based management of these vulnerable coastal habitats.

1. Introduction

Seagrass meadows are highly important and rich habitats on shallow soft sediment bottoms, providing both nature and humans with several crucial ecosystem functions and services [1]. They play a major role in the marine ecology of coastal and estuarine areas worldwide, supporting finfish, shellfish, and invertebrate communities, as well as filtering coastal waters, dissipating wave energy and anchoring sediments [2]. Seagrass meadows are a prime example of a shallow water habitat where both natural (e.g., extreme weather events, overgrazing, and disease) and anthropogenic (e.g., nutrient overload, dredging and filling, pollution, destructive fishing, and anchoring) disturbances can result in patchiness, variable cover, and fragmentation [3]. According to Guiry [4], Nanozostera noltei is currently accepted taxonomically. However, due to notoriety, the former name Zostera noltei Hornemann, 1832 will be used in this manuscript. Z. noltei is one of the predominant global species in intertidal zones, representing the land–sea interface. The distribution of Z. noltei extends from the eastern Atlantic shores starting from Mauritania to southern Norway and the Kattegat Sea and throughout the Mediterranean, Black, Azov, Caspian, and Aral Seas, as well as the Canary Islands. Z. noltei beds serve as a habitat that facilitates the spatial structuring of many biological communities and forms the basis of exceptional trophic compartments for many primary consumers. However, these meadows are particularly vulnerable to climate-change-related effects, such as increasing sea water temperatures. These changes affect the state, composition, and evolution of occupied areas, ranging from homogeneous sectors to fragmented patches scattered along foreshores [5]. The most extensive seagrass meadows in the Black Sea are found in its north-western part and along the Crimean coast (Ukraine and Russia), where they grow in large bays and gulfs, coastal lagoons, and river mouths and deltas. This region contains most of the seagrass habitats in the Black Sea [6]. The Romanian Black Sea coast, which is relatively open and exposed to currents and winter storms, offers limited development areas for Z. noltei. Currently, the main meadows can be found towards the south of the coast, at Mangalia. However, historical data indicate that in the 1970–1980s extensive meadows formed by Zostera marina and Z. noltei were once present at Cape Midia, Eforie South, Tuzla, Comorova, Costinești and Mangalia port [7]. In general, the number of epsilon species diversity in the Black Sea is lower compared to the Mediterranean, a discrepancy attributed to differences in environmental parameters, which influence the ecology and taxonomic composition of bottom vegetation.

The decline of Z. noltei is not a universal feature. This species frequently derives benefit from the decline of other seagrasses and the introduction of restoration plans. In the heavily populated and industrialized Gulf of Thermaikos (Greece), Posidonia oceanica and Cymodocea nodosa losses were followed by the conspicuous expansion of Z. noltei [8]. In the Northern Lagoon of Tunis (Tunisia), after the virtual collapse of Z. noltei due to a dramatic increase in pollution (1920s through 1980s), an ambitious plan for (i) sewage treatment, carried out in 1988, (ii) improvement of water exchange with the open sea, and (iii) establishment of a counter-clockwise water circulation led to the recovery of mixed meadows of Z. noltei, Ruppia sp., and C. nodosa. In sites with fluctuating salinity and turbidity, the extension of Z. noltei follows a consistent pattern of change over the years. For example, in the Vaccarès Lagoon (southern France), Z. noltei declined to almost total absence from 1995 to 1997, in parallel with a decrease in water salinity and an increase of turbidity due to a major flood, then recovered from 1998 to 2000. Z. noltei is sensitive to high salinity, and adverse effects from brine discharges from seawater desalination plants can also be expected [9]. Among the drivers of seagrass decline, eutrophication is the most marked long-term change in the Z. noltei beds, and in general, there is a clear relation between the decline of seagrasses and eutrophication [10]. Eutrophication can also lead to an increase in primary production and to a progressive replacement of Z. noltei meadows by opportunistic species belonging to Gracilaria sp. and tubular forms of Ulva [9]. Also, in a survey of Western European populations of Z. noltei conducted in 1989 and 1990, a few specimens with wasting disease-like damage patterns were found in all investigated populations, but no large-scale deterioration in Z. noltei beds was recorded (OSPAR, 2009). Small-scale decline in Z. noltei was also observed in the Dutch Wadden Sea, where bioturbation caused by an increased density of lugworms (Arenicola marina) smothered young shoots with sediment. This species is highly sensitive to substrate loss, smothering, changes in wave exposure, the introduction of invasive species, and extraction of other species, such as cockles [11].

Although Z. noltei is listed as a species of minor concern by the International Union for the Conservation of Nature (IUCN), the overall population is reported to be declining. Along the Romanian Black Sea coast, this species is Critically Endangered (according to IUCN-like criteria) and is included in the List of Endangered Marine Species (according to Order no. 488/2020 150 published in the Official Gazette no. 300 of 9 April 2020, in Romanian). The increased eutrophication pressure in the western Black Sea in the 1980s and subsequent increase in opportunistic green and red algal blooms resulting in a decrease in water transparency, as well as bottom hypoxia, led to a degradation of phytobenthic communities in both Cystoseira sensu lato beds and Zostera seagrass meadows [12].

In recent years, there has been growing interest in using seagrass restoration as a measure to combat habitat loss, both by attempting to recreate historic habitats and as a compensatory measure when eelgrass is damaged or at risk of disappearing during exploitation or anthropogenic activities (e.g., hydrotechnical constructions). To support future ecological restoration activities, the first step is to provide a baseline study by presenting the specific situation of Z. noltei along the Romanian Black Sea coast. Comprehensive baseline datasets can be used to identify representative metrics of ecosystem health, determine appropriate temporal and spatial scales for monitoring, establish efficient sampling methods, and assess the statistical power required to detect change [13].

Active and passive remote sensing approaches are also used frequently to estimate the coverage and quality of seagrass habitats. Using active acoustic remote sensing methods such as side-scan sonar, it has been shown effective to estimate the densities and coverage of macroalgae and seagrass. Mapping using remote sensing could be used for interpreting preliminary data before field observation, which is also the case in this study. In addition, it can be used as comparison of the field observed data with satellite images and increase credibility of the result of the research [14]. Unmanned aerial vehicles (UAVs) provide an attractive alternative to crewed airborne data capture, especially when repeat captures are required for time series monitoring. UAVs are relatively inexpensive, and the increased accessibility of UAV technology has enabled new imaging methods that can capture fine-scale features even in structurally complex meadows. UAV-acquired images have been used for mapping seagrass density, ecosystem health, and species composition at fine scales, as well as for identifying disturbances such as boat propeller scars [15]. The spectral complexity found in hyper spherical optical remote sensing studies on Z. noltei leaves suggested that the addition of further spectral bands may produce better discrimination between seagrass shoots and background sediments [14].

To enhance understanding of the ecological status of Z. noltei meadows, this study aims to assess and map these seagrass habitats in Mangalia, employing an integrative approach that combines modern techniques, including UAV mapping and ground-truthing, with in situ observations. This methodological combination offers a comprehensive assessment, establishing a reliable baseline essential for tracking changes in seagrass meadow health and resilience to environmental stressors. Furthermore, the study provides valuable insights to support future restoration initiatives by identifying potential donor meadows and offering baseline data to evaluate the success of restoration efforts. Such baseline data are also critical for determining the timescale required for restored habitats to attain comparable ecological functionality to natural meadows. Spatially extensive information, in contrast to field data from point, line, or area samples, is recognized as an essential component to understanding, monitoring, and modelling vegetation species composition, structure, functions, and dynamics in terrestrial and aquatic environments. These data can form the basis for quantifying species richness, abundance, and diversity in a spatial context, as well as assessing the forms and patterns of vegetation patches, and most importantly, are useful for assessing their change over time. Developing effective approaches to map seagrass species, cover, and biomass would provide baseline datasets for these applications [16].

Accurate maps of seagrass properties should provide the basis for developing spatially explicit datasets to support scientific studies for understanding the composition of the environment, along with changes to their spatial structure [16]. These maps would also enable assessment of the controls on seagrass biodiversity, for scaling-up predictive models from site to ecosystem scales. All these spatial datasets may play critical roles in the selection and design of protected seagrass areas, and assessment of the effectiveness of management controls on the biodiversity of seagrass environments [17]. Monitoring and mapping seagrass meadows is of prime importance in establishing the ecological status of such important habitats and is a useful approach to estimating the impacts of natural and anthropogenic stressors [5]. Information on Zostera meadows from the Romanian Black Sea coast is limited, and this integrative study will fill the knowledge gaps on species diversity and distribution. It provides an inventory of Zostera habitats and also a spatial framework for baseline descriptions and monitoring programs.

2. Materials and Methods

2.1. Study Area

Mapping and monitoring of seagrass habitats were carried out by high-resolution orthophotography. Images were obtained in spring–early summer 2023, under calm wind conditions and low levels of glint on the ocean surface. These requirements are quite exact, and the time window to obtain high-quality imagery can be very short [18]. Along the Romanian Black Sea coast, seagrass meadows are distributed in shallow waters from 0 to 3 m depths along non-tidal coasts; therefore, the cold season is not always suitable for these activities, due to strong winds or powerful storms, and in late summer, water turbidity increases along the Romanian Black Sea coast reducing considerably the visibility.

Based on NIMRD’s previous monitoring campaigns and aerial observations, seven potential seagrass meadows were identified in the Mangalia area (Figure 1), hereafter referred to as sites Z1, Z2, Z3, Z4, Z5, Z6, and Z7 (Figure 2). After that, a ground-truthing survey was performed to confirm the presence of the Z. noltei patches. During the survey, Z4 was excluded due to the absence of Z. noltei shoots within this area. All other six meadows are monospecific as far as seagrasses are concerned, formed exclusively of Z. noltei. In Mangalia, Z. noltei grows as dense and sometimes scattered patches at depths between 1 and 3 m.

2.2. UAV Mapping Procedure

Before the assessment activities, a preliminary habitat mapping at Mangalia was performed using an Unmanned Aerial Vehicle (UAV) (DJI AIR2S). The maps were created to assess each meadow’s extent and support further in situ activities (Figure 2). Agisoft Metashape 1.7.6. was used to perform photogrammetric processing of digital images. Maps were generated using QGIS 3.20.3. The flight plan was generated using DJI Pilot 2 software installed on the drone smart remote controller. The flight was made at a 60 m height. For the meadows Z1 to Z6, 279 photos were taken, and for meadow 7, 53 photos. Data from five Ground Control Points (GCPs) were used during data processing in Agisoft Metashape. The ground control points were used in the photogrammetry process in the Agisoft Metashape software to correct the georeferencing performed automatically. Final resolution of the orthomosaic was 3 cm/pixel.

Once the seagrass beds had been delineated through high-resolution photographic analysis derived from UAV observations, the images were processed using geospatial software to enhance spatial resolution and delineate habitat boundaries. This methodology involved orthorectification to correct image distortions and georeferencing to accurately align the images with ground coordinates. Segmentation algorithms and supervised classification were applied to distinguish seagrass meadows from surrounding substrates. These processed outputs were then used to guide the selection of patches for subsequent in-field validation and ground-truthing. The validation and ground-truthing process involved navigating to the identified seagrass patches by boat and having divers examine the seafloor at each patch. The divers’ tasks included confirming the presence of Z. noltei in the locations identified in the UAV mapping and assessing whether the GIS polygons accurately represented the extent of the seagrass habitat. Due to the similar visual signatures of macroalgae and seagrass in aerial photography, divers also distinguished between these two to ensure accurate mapping [18]. UAVs allow large-scale mapping but may miss small-scale variations, so in situ sampling, although time-consuming, offers more precision.

2.3. Field Procedure

In addition to the mapping procedure, in situ data collection was carried out at Mangalia from May to early August 2023 to provide further information on the characteristics of each meadow [19,20,21]. Four main parameters were measured:

• Percentage Cover (%)—the percentage cover of Z. noltei within designated quadrats (0.25 m⁻²) to quantify the density and spatial distribution of the meadow (see Figure 3a).
• Shoot Number—the number of shoots within each quadrat (0.04 m⁻²) to assess the condition of potential donor areas across different meadow sections (see Figure 3b).
• Leaf Length Measurement—the length of seagrass leaves in each quadrat (0.04 m⁻²) to evaluate plant growth and environmental suitability within the habitat (see Figure 3c).
• Voucher Specimen Collection—specimens for ex situ qualitative analysis of epiphytic load to better understand associated flora and potential stress indicators affecting the seagrass.

Photographs including the entire quadrat frame with the label, from an angle as vertical as possible, were taken by the divers.

The Mangalia area was surveyed using transects across each of the seven meadows. A distinct number of transects were established in accordance with the size of the meadow (

Table 1. Number of transects and quadrats inside each Z. noltei meadow.

Mangalia Meadow	Total Number of Transects (T)	Total Number of Quadrats (Q)	Number of Quadrats per Transect
Z1	4	10	T1/3Q
			T2/3Q
			T3/2Q
			T4/2Q
Z2	2	5	T1/2Q
			T2/3Q
Z3	1	2	T1/2Q
Z5	3	13	T1/6Q
			T2/4Q
			T3/3Q
Z6	1	4	T1/4Q
Z7	3	9	T1/3Q
			T2/3Q
			T3/3Q

), spanning across the full width from East to West. When the analysis was done, all conditions were covered (edge/center effect, different depths).

Two types of quadrats were used for the evaluation of each meadow:

• 0.25 m² quadrats (50 cm × 50 cm), placed at regular intervals of 10 m along each transect to visually determine the percentage cover (Figure 4a).
• 0.04 m² quadrats (20 cm × 20 cm), placed at regular intervals of 10 m along each transect to count all shoots per quadrat and to measure the leaf length (Figure 4b).

2.4. In Situ Activities

2.4.1. Percentage Cover (%)

A non-destructive method was employed to visually estimate seagrass cover in the field. Using 50 cm × 50 cm quadrats positioned along cross transects, the percentage cover within each quadrat was assessed following established methodologies [10,11,12]. The procedure was simplified since all the meadows were formed exclusively of Z. noltei; therefore, there was no need to make a dissociation between species. To estimate the percentage cover of Z. noltei inside each meadow, the quadrats (50 cm × 50 cm) were photographed, and their coordinates were recorded for mapping validation [5]. For each quadrat, a visual estimation of percentage cover was performed by assigning it to a scale from 0 to 100%.

2.4.2. Counting Shoots Inside Each Quadrat

This activity was performed in situ, using a non-destructive approach, by carefully counting all shoots within each 20 cm × 20 cm quadrat positioned along each transect [10,11,12]. For accuracy and very precise measurements, quadratic reporting was used. This is important to accurately determine which would be the potential donor area for ecological restoration activities. For standardization, only the mean value of the number of shoots calculated for all quadrats in each meadow was reported per square meter by using the multiplication coefficient 25.

2.4.3. Leaf Length

Typically, Z. noltei exhibits seasonal variability in leaf growth, with longer leaves during summer replaced by shorter, slower-growing leaves in winter. To measure the leaf blade length, 5 to 7 shoots were randomly selected within each 20 cm × 20 cm quadrat. Emphasis was placed on selecting shoots where the leaf tips remained intact, excluding blades that were shortened or damaged. Divers carefully extended the cluster of leaves from each shoot to their maximum length without disturbing their roots. Using a ruler, they measured the length from the substrate to the leaf tip [10,11,12]. Measurements were expressed in cm.

2.4.4. Voucher Specimens

Representative specimens (mature shoots with intact leaves, roots, and rhizomes) were randomly collected by hand from each meadow, placed in a Ziplock bag, labelled, and stored on ice [12]. Before analysis, samples were rinsed with distilled water to remove debris, and leaves and roots were spread on clean paper sheets to make each part distinct. First, the shoots were visually analyzed, and then they were submitted to a microscopic analysis using an OLYMPUS SZX10 stereomicroscope (magnification 2× and 4×; OLYMPUS, Feasterville, PA, USA) and an OLYMPUS IX51 inverted microscope (magnification 10× and 20×). Ex situ measurements were also performed to compare the results with those in the field. The longest blade length from each shoot was measured, and the results were recorded in centimeters (cm). The epiphytic load was qualitatively analyzed in terms of diatoms and macrophyte species composition, providing information regarding species type, dominant species, and degree of coverage. Photos were used for validation and to observe the species in its natural environment. The epiphytic load was also estimated through photo analysis.

2.5. Data Analysis

Statistical analysis of data was performed using PRIMER 7 (v.7.0.23) [22] and XLSTAT 2023.1.6. [23]. Statistical analysis was performed on the entire dataset (Z1–Z7, all data) to test differences between meadows. The variables considered for the statistical analysis were the following: number of shoots per quadrat, leaf length (cm), and percentage cover (%). Normal distribution and homogeneity of variances were established before each statistical analysis using the Shapiro–Wilk Normality Test and Levene’s test. The significance level was set at p < 0.05. When normality and homogeneity criteria were not met, a non-parametric Kruskal–Wallis combined with a multiple pairwise comparison (Dunn’s post-hoc test) was selected to test the difference between all meadows. A Demsar plot was used to highlight how the meadows cluster according to each parameter.

3. Results

Shapiro–Wilk’s test showed that the number of shoots per quadrat (W = 0.804, p < 0.0001), leaf length (W = 0.975, p = 0.001), and percentage cover (%) (W = 0.440, p < 0.0001) significantly deviate from a normal distribution among all meadows. Levene’s test showed that data differed markedly among all meadows (F (2, 300) = 3.026, p < 0.0001).

Non-parametric Kruskal–Wallis revealed statistically significant differences between meadows when considering the number of shoots per quadrat (p < 0.0001) and leaf length (cm) (p < 0.0001). At the same time, no statistically significant differences were observed when considering Z. noltei percentage cover (%) (p > 0.05).

Z1 had significantly higher shoot densities (a mean of 49 shoots/quadrat⁻¹) compared to all other meadows, except for Z7 (with a mean of 24 shoots/quadrat⁻¹). The lowest mean values were recorded for Z5 and Z6 (15 shoots/quadrat⁻¹) (

Table 2. Summary statistics for all Z. noltei meadows (Z1–Z7).

Meadow	Minimum	Maximum	Mean	Std. Deviation
Number of shoots/quadrat−1
Z1	26	74	49	15
Z2	16	26	20	4
Z3	18	18	18	0
Z5	7	24	15	4
Z6	12	17	15	2
Z7	17	32	24	5
Leaf length (cm)
Z1	8	54	36	9
Z2	30	54	43	6
Z3	34	75	53	17
Z5	5	99	60	27
Z6	47	61	54	5
Z7	18	99	60	19
Percentage cover (%)
Z1	75	100	95	11
Z2	90	100	94	5
Z3	100	100	100	0
Z5	90	100	98	4
Z6	100	100	100	0
Z7	100	100	100	0

). Dunn’s post-hoc test showed statistically significant differences between Z1 and all other meadows, except Z7. Other statistically significant differences were observed between Z5 and Z7 (p = 0.002) and Z6 and Z7 (p = 0.026) (Figure 5).

The shortest leaves were measured inside Z1, with an overall average leaf length of 36 cm, while the longest were observed across Z5 and Z7, with an overall average leaf length of 60 cm (Table 2). Nevertheless, we would like to emphasize that this is not necessarily a sign of the inferior ecological quality of Z1. Because there were no variations in depth between polygons, we assumed that the differences were related to monitoring campaigns periods; Z1 was monitored at the end of May, the other meadows at the end of July. Therefore, it is normal for the leaves to increase in length during the two summer months, hence the differences between the measurements. Regarding leaf length (cm), statistically significant differences were observed between Z1 and Z5 (p = 0.004), Z1 and Z6 (p = 0.030), and Z1 and Z7 (p = 0.003) (Figure 6).

According to the in situ observation, shoots of Z. noltei were visually identified along all transects. The meadows are considered to be compact, with an overall coverage of more than 90% (Table 2).

The Demsar graphs showed that the factor differentiating meadows the most was the number of shoots. Considering this, the meadows clustered into three groups (Figure 7a). For leaf length, the similarity between meadows was also high, whereas only two clusters were distinguished (Figure 7b). The highest similarity was in the case of coverage, as all meadows clustered into one group (Figure 7c).

Mean shoot density varied between a minimum of 365 shoots/m⁻² inside Z5 and a maximum of 1223 shoots/m⁻² inside Z1 (Figure 8). A slightly negative correlation between number of shoots and average leaf length was observed across all meadows (Figure 9).

To have a visual image of the current situation Z. noltei meadows, maps were produced based on in situ measurements (i.e., number of shoots, average leaf length, percent coverage) for each of the six meadows (Figure 10, Figure 11 and Figure 12). The maps provide a visual representation of the shape of each meadow, which allows an evaluation of whether a vegetated area is likely to become more fragmented.

The epiphytic load was similar between the meadows, with no notable differences in terms of its qualitative aspect. Generally, the epiphytic load consisted mostly of green and red macroalgae and benthic diatoms. Green algae were represented by the opportunistic Ulva rigida, U. intestinalis, Cladophora spp., and Chaetomorpha spp. As for the red algae, the following were observed on the blades: Acrochaetium spp., Callithamnion corymbosum, Ceramium virgatum, and Carradoriella denudata. Among diatoms, the main species reported were Navicula spp., Nitzschia sp., Synedra spp. Licmophora spp., Striatella delicatula (Figure 13a), and Achnanthes longipes. Diatoms’ cells were strongly adherent to the leaves, forming abundant epiphytic colonies.

Of all epiphytes, the dominant was Acrochaetium spp., a small red alga known to usually have an abundant development during the summer season on Z. noltei blades, providing a ‘’furry” appearance (Figure 13b). In addition to these common epiphytic species, it is worth mentioning the presence of a species not so common on the Romanian Black Sea coast, the small brown algae Myrionema orbiculare, found as an epiphyte on Z. noltei collected from Z3, Z5, and Z7 meadows (Figure 13c).

4. Discussion

The global loss of seagrass from coastal waters is a well-documented phenomenon [24]. On the Romanian Black Sea coast, the decline of seagrass habitats has been noticeable, starting in the 1980s. This decline has been attributed to eutrophication, with the Romanian coastline disproportionately impacted by nutrient input from the Danube, driven primarily by human activities such as agricultural runoff, industrial discharge, and urbanization of the river basin, which led to a significant increase in eutrophication in the Black Sea during the 1980s and 1990s [25]. In recent years, eutrophication has decreased, and coastal ecosystems are showing signs of recovery. However, ongoing coastal reconstruction projects continue to pose significant threats to this vital habitat-forming species. To mitigate the impact, the provision of new, larger areas with appropriate conditions for the implantation of Z. noltei is included as part of the coastal rehabilitation works. Implantation of Z. noltei into some of these areas in nearby coastal rehabilitation sites has been, and is planned to be, carried out by relocation of sods of this species from the study site’s affected areas to other favorable areas.

In our study, the restoration activities were undertaken as a direct response to beach nourishment projects conducted along the Romanian shoreline. These nourishment activities will inevitably result in all Zostera meadows being destroyed in the upcoming years. To mitigate the loss, ecological reconstruction activities in nourished sites were foreseen. The selected relocation site had already undergone nourishment, providing an environment deemed suitable for the growth of Z. noltei based on local ecological assessments. Consequently, the Z. noltei from the area that will be affected was successfully relocated to this area, ensuring the preservation and continuity of this vital habitat-forming species. It must be emphasized that restoration activities necessitate meticulous preparation. This baseline study served as a foundational step preceding the relocation process.

Active restoration involves transplanting Z. noltei seeds, seedlings, or mature plants into degraded sites. These transplants are typically scheduled to coincide with seasonal growth cycles and are tailored to local environmental conditions to maximize success. For example, restoration projects in France and the Netherlands also use re-seeding techniques that collect seeds from nearby donor meadows, which helps maintain genetic diversity and enhances adaptive potential within restored populations.

Passive restoration, in contrast, focuses on improving environmental conditions to support natural recovery. This includes measures such as reducing water pollution, establishing no-anchor zones to prevent physical damage, and implementing conservation regulations in sensitive areas.

To support active and passive restoration initiatives and provide a reference for measuring progress, establishing baseline assessments of Z. noltei meadows is essential. This includes conducting detailed surveys on the current distribution, health, metric characteristics, and environmental conditions of Z. noltei in the region.

We successfully employed UAVs for aerial monitoring and SCUBA divers for ground-truthing to evaluate seagrass meadows in the Mangalia area, Romania. This study marks the first instance of this integrated approach being utilized in Romania, representing the most comprehensive assessment of Z. noltei meadows along the Romanian Black Sea coast to date. In 2008, Surugiu noted that Zostera beds disappeared from the Romanian coastline due to eutrophication and pollution in the 1980s. No further studies on this species were published until a small Z. noltei bed of approximately 20 m² was reported in Mangalia at that time. The article only mentions the presence of this meadow in Mangalia and presents some ecological characteristics of this species [26]. Dinu et al. [27] highlighted that there is no information available on Z. noltei stem density along the Romanian Black Sea coast. This accurate information would have been extremely useful in modeling simulations carried out to provide a quantitative assessment of the effect of a Z. noltei meadow on coastal protection for an area located on the south coast of Romania. By mapping all meadows from Mangalia and providing information on shoot density, leaf length and coverage, this study filled this knowledge gap.

By integrating both in situ and ex situ innovative methodologies, our research provides a critical baseline for current and future restoration efforts. This integrative monitoring approach deepens our understanding of the species’ status (i.e., number of shoots, leaf length, epiphyte load), facilitates the identification of ecological threats, and guides the development of locally tailored restoration strategies, such as selecting suitable donor sites and assessing biological material availability for translocation. Given Z. noltei’s protected status and sensitivity to both anthropogenic and natural disturbances, non-destructive methods were prioritized to minimize impacts on this vulnerable species.

Seagrasses are considered ecosystem engineers due to their capacity to modify their environments, thereby creating distinctive habitats. Along the Romanian coastline, Z. noltei meadow habitats are protected and classified under Natura 2000 (https://natura2000.eea.europa.eu/) as 1110-1 Fine clean or slightly muddy fine sands with Zostera meadows. Additionally, according to the EUNIS classification (https://eunis.eea.europa.eu/habitats.jsp), these meadow habitats are categorized as Black Sea seagrass meadows situated on moderately exposed upper infralittoral clean sands (MB547).

It was previously known that the main Z. noltei meadows in Romania are located along the southern coastline, specifically in the Mangalia region. A few patches of Z. noltei exist in Navodari and 2 Mai also. However, their ecological characteristics are notably less developed compared to those in Mangalia. Additionally, we report here for the first time the discovery of a small Z. noltei meadow within Mangalia Harbor, identified during this year’s survey, which requires further assessment to evaluate its ecological and biological features. Our habitat mapping was in line with previous knowledge, confirming dense Z. noltei meadows on sandy substrates in the Mangalia coastal area.

Zhang et al. [2] support that shoot density, dimensions, and weight of seagrass in Swan Lake, China exhibited strong seasonal variation, with maximal values in summer and minimal values in winter and early spring. A study from two Mediterranean lagoons also revealed that temperature and turbidity are significant environmental factors influencing the temporal changes in Z. noltei meadows [28]. In general, differences were most pronounced for aboveground plant components (i.e., shoot density and leaf length). Berov et al. [12] mention stem densities for Z. noltei between 750 and 2250 stems/m⁻² and stem heights between 20 and 40 cm in the area of Burgas Bay (Bulgaria). In the case of Z. noltei from the Romanian coastal areas, the density is much lower (Figure 8), but also the ecological conditions are less favorable, with no sheltered areas, compared to Sevastopol Bay and Burgas Bay. In contrast, shoots displayed much longer leaves than those reported for the Bulgarian coast, with stem heights ranging from 36 to 60 cm (Table 2). As the Zostera habitat is distributed in shallow waters along the Romanian coast, the negative correlation between shoot number and average leaf length could be a consequence of the fact that lower density may allow light to penetrate more easily, thus supporting faster growth of leaf length. Shoot density responds rapidly to disturbances such as shading [29]. However, this is a dynamic aspect, and the analyzed variables may show significant seasonal variations.

For the data presented in this study, generally, the greatest differences were observed between Z1 and all other meadows (Table 2). The notable variations in measured parameters at Z1 could also be attributed to the timing of data collection, which occurred in late spring, as water temperatures began to rise, and plants entered a phase of active growth and regeneration following winter disturbances. Wong et al. [13] observed distinct seasonal patterns in the morphometric parameters of seagrass, noting that shoot density, dimensions, and biomass tend to reach peak values in summer and decline to their lowest in winter and early spring.

Determining the status and health of naturally occurring meadows can help inform decisions on which meadows are most suitable to act as donor populations. For example, meadows that have a high density, healthy leaves, and high percentage cover are more likely to provide stronger, more resistant material for restoration sites and are also more likely to recover quickly from the physical disturbance of harvesting transplant material. In our study, Z1 fulfils these requirements and was chosen as a donor area for restoration activities. To date, harvesting of biological material for restoration purposes was conducted in 2023, during which Z. noltei sods were transplanted to the Eforie area. These relocated sods are currently under monitoring to assess their establishment and adaptation to the new environment. Coverage of Z. noltei in temperate latitudes usually displays unimodal seasonal curves, with a maximum in summer and a minimum in winter. Our study showed a mean percentage coverage ranging between 94% and almost 100% in all sampling sites.

Regarding epiphytic flora, a proliferation of Ulva sp. might result in a reduction in light availability and probably interfere with the intrinsic temporal variation of biomass, growth, and carbohydrate reserves of the seagrass itself [17]. Eutrophication-gained filamentous algae (mainly ephemeral) may shade seagrasses, hamper water exchange, and cause a decline in associated fauna [30]. Seagrasses with excessive leaf epiphyte cover are thus more prone to anthropogenic impacts and activity in coastal environments. Leaves with an abundant fouling reduce their resilience to environmental disturbances [31]. However, the absence of epibionts is not a sign of good health since it is well known that seagrass meadows are important habitats, supporting a rich biodiversity. Epiphytic blooms may vary markedly over time, both because they grow fast and because they are regulated by wind exposure and can be decimated after a storm [30]. Filamentous brown and green algal mats were observed fouling seagrass, the sea bottom, and the water surface in the spring and summer. Many studies have shown that fouling of these plants by filamentous algal mats reduces light availability, with negative consequences for growth and production [13]. During the summer of 2023, no green algae blooms were observed near these seagrass meadows, which allowed a normal development of the shoots. In general, the leaves presented a light to moderate epiphytic load, meaning that one to two thirds of the blade had epiphytes. The epiphytic species were mainly present in the apical part of the leaves. In Z. noltei meadows along the Romanian coast, the epiphytic load observed during this study aligns with typical levels documented annually during the summer season in monitoring surveys. At this level, epiphytic presence appears to exert minimal impact on the seagrass meadows and does not significantly affect seagrass health or function. Ex situ analysis showed that the seagrass shoots randomly harvested from each meadow were green, with no signs of blackening (an indicator of poor health) and with minimal or no blade damage.

5. Conclusions

This study establishes critical baseline data on Z. noltei meadows along Romania’s Black Sea coast. The comprehensive baseline assessment of Z. noltei meadows near Mangalia provided essential reference data for conservation actions, specifically supporting the selection of site Z1 as a reference donor for seagrass restoration along the Romanian coast in 2023. By documenting the structural and ecological health of these meadows, this study offers a robust foundation for future restoration initiatives.

In situ and ex situ analyses revealed generally healthy meadow conditions, with moderate epiphytic fouling, characteristic for the summer season. This indicates resilience in these meadows despite environmental pressures and provides insights into current health, which can inform ongoing management and mitigation of future anthropogenic impacts. UAV mapping and high-resolution orthophotography have proven effective for detailed habitat assessment, enabling spatially explicit monitoring essential for long-term tracking of meadow health.

The dataset and all the information provided in this manuscript can be extrapolated to other areas and can serve as a European-wide comparison between different Zostera meadows. This will help create an information network on the ecological status of different Z. noltei meadows in Europe.